Kathy Garcia
@NeuroKathyG
Followers
135
Following
114
Media
5
Statuses
35
Computational Cognitive Science PhD @JHUCogSci | BS @Stanford |
Joined June 2020
Excited to be presenting this paper at #ICLR2025 this week @iclr_conf! Come check our poster if you want to know more!! 🧠🤖 . 📆 Thur, Apr, 24 3:00-5:30 - Poster session 2 (#64) .📄
openreview.net
Deep learning models trained on computer vision tasks are widely considered the most successful models of human vision to date. The majority of work that supports this idea evaluates how accurately...
0
4
11
RT @nmblauch: What shapes the topography of high-level visual cortex?. Excited to share a new pre-print addressing this question with conne….
0
30
0
RT @RobinNusslock: My lab at Northwestern is hiring! We have openings for a Project Coordinator, Research Assistant, MRI Data Analyst, and….
0
42
0
RT @genex_world: Introducing GenEx: Turn any image into a 3D world adventure!. 1️⃣ Create a fully explorable 360° world in 3D from just a s….
0
12
0
RT @paulsoulos: 🚨 Thrilled to share that Compositional Generalization Across Distributional Shifts with Sparse Tree Operations received a s….
0
4
0
RT @lforlasagna: One of the highlights of the Fall semester for me is organizing our PhD application mentorship program!!🧑🎓 We offer SOP a….
0
6
0
RT @luciedqt: Our kidney work being finally out, that means I get to share some of my favorite images 🤩 Starting off with the results of ou….
0
7
0
RT @khai_loong_aw: Accepted to @COLM_conf! 🎉. Instruction-tuning helps LLMs produce more human-like responses. Does it also make LLMs 🤖 mor….
0
10
0
RT @_LXAI: Don't miss our LatinX in AI Workshop at ICML! 🌟 It's just around the corner! Connect with our incredible keynote speakers and be….
0
7
0
RT @nmblauch: Missed out on ICLR Re-Align? No fear - here is a 🧵 on our new work: "Topoformer: brain-like topographic organization in Trans….
0
26
0
RT @emaliemcmahon: I had so much fun working on this with Colin, @michaelfbonner and @leylaisi, but most of all @NeuroKathyG. It was a joy….
0
1
0
🚀 Together, this work highlights a major gap in AI's ability to match human social vision, and underscores the importance of developing AI models in dynamic social contexts [6/7].
1
0
3
📹 While most model features (like architecture or training objective) did not affect performance, we saw a big advantage for video versus image models along the lateral stream. but no model tested could predict lateral stream responses well. [5/7]
1
0
5
🔍 Unlike visual scene features and ventral stream responses, vision models struggled to match human action and social interaction ratings, and did a poor job of predicting brain responses along the recently proposed lateral stream, specialized for social perception. [4/7]
1
0
6
🧠 We benchmarked 350+ image, video, and language models against human behavioral and neural responses to dynamic, social videos. [3/7]
1
0
5
🎥 Real-world vision is dynamic, involving complex social interactions. Current AI models provide a good match to humans in static scene vision, but how do they fare with dynamic, social stimuli? 🤔 We set out to explore this! [2/7]
1
1
5
📢 Excited to announce our latest pre-print: “Modeling dynamic social vision highlights gaps between deep learning and humans”! 🌟 w/ @emaliemcmahon, Colin Conwell, @michaelfbonner, and @leylaisi🧵 #NeuroAI #CognitiveScience [1/7]
3
19
47